Pulp & Paper Mill Transformer: How to Size the MV Distribution for a Paper Machine Line, Why VFD Harmonics and Refiner Starting Decide the Spec, and When Dry-Type Beats Oil

The answer first: a pulp and paper mill transformer is sized from the area load list and from the ride-through requirement of the paper machine, not from a kWh-per-tonne benchmark. What actually decides the rating is three things that have nothing to do with average demand: the harmonic current that the sectional drives of the paper machine push back into the 690 V bus, the starting current of the largest refiner motor in the pulp mill, and the fact that a 150 ms voltage dip costs the mill a paper break while a 300 ms dip costs it a restart. A competent specification therefore starts with a load list by mill area, applies demand and diversity per area, derates the transformer for the measured harmonic spectrum, checks the voltage dip at the largest motor start, and only then fixes the impedance that keeps the 690 V and 11 kV switchgear inside their fault ratings.

Key takeaways

• A modern paper machine is a drive system before it is a load: 80–150 sectional drives on one 690 V bus, most of them regenerative, most of them with a diode or active front end that decides the transformer’s harmonic duty.
• Size the paper machine transformers from the area load list with a 0.5–0.75 demand factor, then derate for harmonics with IEEE C57.110 — a 690 V bus with 6-pulse front ends loses 13–20% of its transformer capacity before any margin is added.
• The largest single motor in a pulp mill — a 10–30 MW TMP refiner — sets the minimum bus strength. A 12 MW refiner started direct-on-line on a 40 MVA, 8% transformer pulls the 11 kV bus down 12.5%; on 25 MVA it pulls it down 18.7% and trips every drive on the same board.
• Dry-type cast-resin transformers belong inside the machine hall, the stock preparation area and the bleach plant; oil-immersed transformers belong outdoors at 33 kV and on the generator step-up. The decision is fire load and corrosion, not price.
• For a mill running 8,000 hours a year, 1 kW of no-load loss is worth about USD 7,300 of capitalised cost over 20 years at 0.08 USD/kWh — more than the price difference between efficiency grades on a 4 MVA unit.

Why is a paper mill an electrical problem before it is a transformer problem?

An integrated paper mill is not one load with one load factor. It is eight or nine process areas with completely different electrical behaviour, strung along 1–2 km of site and tied to a single 132 kV or 110 kV interconnection. The woodyard starts and stops conveyors all day. The digester runs continuous pumps at a flat 90% load. The bleach plant has dozens of small drives in a chlorine dioxide atmosphere. The stock preparation refiners draw 20–40 MW and start maybe twice a day. The paper machine runs 24 hours at 95% availability with a load profile so flat that its demand factor looks like a constant. The recovery boiler and the turbine generator make the site a power producer for part of the year.

Each of those areas has its own transformer, its own fault level and its own reason to fail. That is why the useful unit of specification in a mill is the area substation, not the plant total. A mill-wide kVA number tells you nothing about whether the 690 V bus will survive a coater blade load step or whether the 11 kV board will hold when the refiner starts.

Two consequences follow from this table. First, the mill’s total transformer capacity is typically 1.4–1.8 times its maximum demand, because the areas do not peak together and because every area needs a maintenance outage path that does not stop the machine. Second, the transformer population is mixed: a mill of 700 t/d will have 25–40 transformers above 1 MVA, and the ones that cause downtime are almost never the biggest.

How much connected load does a 700 t/d paper machine line actually carry?

Take a coated fine paper machine with an 8.4 m trim, 1,500 m/min design speed and 700 t/d output, which is 29.2 t/h. The machine itself — not the pulp mill behind it — carries about 17 MW of connected load. The load list below is the kind of list that should be attached to the transformer enquiry; if a supplier quotes a paper machine transformer without one, the number is a guess.

At a displacement power factor of 0.96 the demand is 10.9 MVA. That is the number most people stop at, and it is the wrong number. The transformer that feeds this bus sees a distorted current, not a sinusoidal one, so its usable capacity is lower than its nameplate — the arithmetic in the harmonic section below brings 10.9 MVA up to 12.8 MVA. Four 4 MVA units then sit at 80% loading each, which is the loading a paper machine transformer should be designed for: high enough to justify the asset, low enough to absorb a section upgrade or a drive retrofit without a new transformer. Whether 80% is a safe continuous loading for a specific design depends on the hot-spot rise and the daily load cycle, which is the subject of the load factor and overload guide.

Note on units: the paper machine bus is usually specified at 690 V, not 400 V. At 690 V a 4 MVA transformer delivers 3,347 A, which is inside the range of standard 690 V switchgear. At 400 V the same 4 MVA would need 5,774 A, which forces parallel cables, parallel switchgear and a much larger fault current at the MCC. That single choice — 690 V against 400 V — is the most consequential voltage decision on a paper machine, and it is usually made by the drive vendor before the transformer is discussed.

What voltage architecture should a paper mill use?

Most mills land on four levels, with the exact numbers set by the local utility rather than by preference. In Europe and much of Asia the chain is 132 or 110 kV in, 33 or 35 kV around the site, 11 or 10 kV to large motors and area substations, 690 V to drive systems and 400 V to lighting and small power. In North America the same architecture appears as 138/115 kV, 34.5/13.8 kV, 4.16 kV and 480 V.

The voltage level is not a free choice: the 33 kV ring exists because the mill is physically long and 11 kV cables to the far end of the site would be uneconomic at 20 MW. The 11 kV level exists because 8–30 MW motors are not built at 690 V. The 690 V level exists because drive power ratings taper off above about 2 MW and because the paper machine needs hundreds of small drives on a common bus. Each level has a transformer between it and the next, and each of those transformers has a different job: the 132/33 kV units carry the site, the 33/11 kV units carry starting current, the 33/0.69 kV units carry harmonics. The same reasoning, applied to the 6, 10, 11, 20 and 33 kV options individually, is set out in the voltage class selection guide.

Why is the 690 V drive bus the real centre of a modern paper machine?

A paper machine is a continuous web running at 1,500–2,000 m/min through a sequence of nips: headbox, wire, press, dryer, size press, calender, reel. The web cannot tolerate a speed mismatch between sections, so every section is driven by its own motor under closed-loop tension or speed control. A modern machine has 80–150 such sections, from 5 kW for a wire return roll to 400–600 kW for the reel drum.

That architecture produces a load with properties that a conventional distribution transformer was not designed for: near-unity displacement power factor, a high harmonic content from the rectifier front ends, and — on unwind stands, winders and calenders — sustained reverse power flow when the drive decelerates or holds tension. The transformer on that bus is therefore not a passive kVA device. It is part of the drive system, and its impedance, its harmonic capability and its thermal time constant all show up in the machine’s behaviour.

The common DC bus is what changed the arithmetic. When 40 sections share a DC bus, the energy released by a decelerating dryer section feeds a section that is accelerating, and only the net difference goes back to the AC side. That reduces the number of braking resistors, reduces the thermal cycling of the transformer, and makes the AFE supply unit — typically 2–4 MW — the only rectifier the transformer actually sees. The harmonic problem does not disappear, it concentrates: instead of 100 distributed 6-pulse rectifiers, you have two or three large AFE units, and the transformer is sized around those.

How do you size the transformer behind a 690 V paper machine drive bus?

The calculation has four steps and each one can change the answer. Using the load list above:

• Step 1 — demand kVA: 10,475 kW ÷ 0.96 = 10.9 MVA. This assumes the drives correct displacement power factor to near unity, which AFE drives do.
• Step 2 — harmonic derating: if the bus has 6-pulse front ends at THDi 35%, IEEE C57.110 derating at K-factor 4.7 allows 85% of nameplate, so the required nameplate is 10.9 ÷ 0.85 = 12.8 MVA. With AFE drives at THDi 4% the derating is negligible and the requirement stays at 11.0 MVA.
• Step 3 — standard rating and split: 12.8 MVA rounds to 4 × 4 MVA on two bus sections, wet end and dry end, with a bus tie that is normally open. Four units rather than two means a single transformer failure takes out half a machine, not the machine.
• Step 4 — checks: voltage regulation at the drive terminals must stay inside ±10% of 690 V in steady state and above the drive’s undervoltage trip during a section acceleration; short-circuit current at the 690 V terminals must stay inside the switchgear rating.

Step 3 is where paper machines differ from general distribution. A 4 MVA, 690 V transformer with 8% impedance has a 50 MVA short-circuit capacity at its terminals, which is 41.8 kA — inside a 50 kA switchgear rating but not inside a 40 kA one. Dropping the impedance to 6% raises that to 55.8 kA and forces 65 kA switchgear. Going the other way, to 10%, cuts the fault current to 33.4 kA but adds 1.5–2% to the drive-terminal voltage drop at full load and makes the AFE work harder to hold the DC bus. On a 690 V bus, 6–8% is the practical window and 8% is the safe default. If you need the fault current at the MCC rather than at the transformer terminals, the three-step method in the short-circuit current calculation includes the cable and motor contribution terms.

Voltage drop figures are the transformer’s own regulation at rated current, computed from impedance and X/R ratio; they exclude cable drop and upstream impedance. On a real 690 V bus the cable drop from transformer to MCC can add another 1.5–2.5% on long runs, which is why the transformer is usually placed within 50 m of the drive line-up.

Why do VFD harmonics cost more transformer capacity than the load list suggests?

A transformer rated 4 MVA at 50 Hz with a sinusoidal load is rated 4 MVA at 50 Hz with a sinusoidal load. Put a 6-pulse rectifier on its secondary and the same 4 MVA of apparent power now contains 35% harmonic current, and those harmonics do two things the fundamental does not: they increase the rms current for the same real power, and they concentrate in the windings through skin and proximity effect, so the eddy-current loss rises roughly with the square of the harmonic order. IEEE C57.110 exists precisely because the second effect is not captured by the first. The mechanism, and the five ways to reduce the harmonic current at source rather than at the transformer, are covered in the transformer harmonics guide.

The derating is computed from the harmonic spectrum, not from a rule of thumb:

I_max (pu) = √[ (1 + P_EC-R) / (1 + K · P_EC-R) ]

where P_EC-R is the eddy-current loss fraction of the total load loss at rated current (typically 0.10–0.15 for a distribution transformer) and K is the harmonic loss factor, K = Σ(h² · I_h²) / Σ(I_h²). The K-factor is dominated by the low-order harmonics because of the h² weighting: the 5th harmonic contributes 25 times as much per unit as the fundamental, which is why a spectrum with 30% fifth harmonic produces K ≈ 4.7 rather than the K ≈ 1.2 that a naive reading of THD would suggest.

Two practical conclusions follow. First, a K-rated transformer does not give you more capacity — it gives you a transformer that survives the K-factor you actually have, with a neutral and a winding design that will not overheat. The capacity still comes off the nameplate. Second, the cheapest way to recover the lost 13–19% is to specify AFE drives on the paper machine in the first place: the derating disappears, the regeneration comes free, and the 33 kV bus stops needing a detuned filter bank. Compare that against the price of one transformer size up on every machine section before deciding that the 6-pulse drives are the economical choice. The decision framework for K-rated versus oversized is set out in the K-factor transformer selection guide.

What is the largest motor-start event in a pulp mill — and how much voltage dip does it cause?

It is not the paper machine, which starts on drives and ramps. It is the refiner. A thermomechanical pulp line has two or three refiners, each driven by a 10–30 MW synchronous or induction motor at 11 kV, and each start is a system event: the motor draws 5–6 times its rated current for 10–40 seconds while the refining plates come up to speed against a chip load.

Run the numbers for a 12 MW refiner motor, 11 kV, 0.96 efficiency, 0.95 power factor, locked-rotor current 5.5 times rated:

• Motor rated apparent power = 12,000 ÷ (0.96 × 0.95) = 13.2 MVA
• Starting apparent power = 5.5 × 13.2 = 72.4 MVA
• At 11 kV, that is a starting impedance of 11² ÷ 72.4 = 1.67 Ω at a starting power factor of about 0.25

The dip at the 11 kV bus then depends almost entirely on how strong the source is. On a 40 MVA transformer with 8% impedance the source impedance is 0.242 Ω and the bus falls to 87.5% of nominal — a 12.5% dip. On a 25 MVA transformer the same start produces 18.7%, and on a 63 MVA transformer 8.3%.

The design rule most mills apply is a 10% dip limit at the 11 kV board and 15% at the motor terminals, because contactors drop out around 70–75% and drives trip on their own undervoltage settings somewhere between 80% and 85%. A 12.5% dip on a 40 MVA board is therefore not a marginal case — it is above the limit, and it will drop every contactor on the same section that is not on a UPS-backed control circuit. That is the moment when the mill discovers that the refiner start and the paper machine share a bus section.

The starting power factor matters more than most load-flow studies assume. At a starting power factor of 0.25, almost all of the motor’s starting impedance is reactive, so the dip is an X/R problem rather than an R problem, and adding cable length to the refiner makes it slightly worse rather than better. This is why a refiner fed from a distant 11 kV substation can show a bigger dip than the same refiner fed from the board next door, even though both are on the same transformer.

How does refiner starting decide transformer impedance?

Here the requirements pull in opposite directions. A higher impedance limits the fault current at the 11 kV board and lets you keep cheaper switchgear; it also increases the dip during a refiner start, increases the transformer’s load loss, and worsens voltage regulation for the whole area. The table below holds the transformer at 20 MVA and varies the impedance for the same 12 MW direct-on-line start.

Read the table the other way round and the design becomes clear: impedance is not the tool you use to make a refiner start acceptable. Even at 6% the dip is 17.7%, because the motor’s own starting impedance dominates. The tool that fixes a refiner start is the starting method — a VFD, an LCI or a pony motor — and once the start is current-limited to 110%, the transformer impedance can be chosen for fault current and regulation alone. Mills that try to solve a refiner start with a bigger transformer end up buying 63 MVA to serve 20 MW of load, which is an expensive way to avoid a drive.

Why does a 100 ms voltage dip break paper — and what does a break cost?

A paper machine has no inertia in the web. The sheet is held in tension between sections, and if one section loses torque for 100 ms while its neighbours keep running, the tension transient either snaps the sheet or produces a thicker, weaker place that snaps at the next nip. The drives themselves usually survive a dip; the sheet does not. The sequence is: dip, one or two sections trip on undervoltage or overcurrent, sheet breaks, broke fills the pulper, and the crew spends 15–45 minutes rethreading from the headbox to the reel.

The cost of a break is not the electricity. On a 700 t/d machine running coated fine paper, a break costs 3–5 hours of production including rethread and broke handling, which at 29 t/h and a contribution margin of USD 300–600 per tonne is USD 26,000–87,000 per event, plus the value of the broke and the downstream schedule disruption. Mills with 10–30 breaks a year therefore spend real money on ride-through, and the transformer specification is part of that spend: a transformer with a lower impedance and a stiffer bus gives the drives more headroom, and a mill with a fast bus transfer scheme can move the machine to the healthy incomer in 200–300 ms instead of losing the sheet.

Oil-immersed or dry-type — where does each belong in a paper mill?

In a mill the choice is driven by location, not by preference. Dry-type cast-resin transformers are the default inside buildings because they have no oil to burn, no bund, no fire wall and no nitrogen or conservator to maintain. Oil-immersed transformers are the default outdoors at 33 kV and above because they are cheaper per MVA, more efficient, and better at dissipating heat in a hot climate.

Two mill-specific points are often missed. First, a cast-resin transformer in a paper machine hall is running in a 40–45 °C ambient with paper dust settling on the windings, so its rating must be corrected for ambient and its enclosure must be designed so that the dust does not bridge the insulation — IP54 with a filtered ventilation path, not IP21. Second, the fire strategy has to be decided before the transformer is ordered: FM Global Data Sheet 5-4 and the local insurer will set separation distances and fire ratings for oil-filled units, and retrofitting a fire wall around an installed transformer is far more expensive than choosing dry-type from the start.

What do the bleach plant, the wet end and the machine hall do to the specification?

Paper mills are wet, warm and corrosive in a way that general industry is not. Four environments drive specification:

• Bleach plant: chlorine dioxide and chlorine atmospheres at 40–50 °C with high humidity. Carbon steel enclosures corrode within 3–5 years. Specify ISO 12944 C5-M coating, 316 stainless hardware, sealed cast-resin windings, IP55, and a corrosion allowance on the enclosure rather than a paint system alone.
• Paper machine hall: 40–45 °C ambient, 80–90% relative humidity at the wet end, paper dust everywhere. Derate for ambient above 40 °C, specify anti-condensation heaters that are energised when the transformer is off, and place the unit so that cleaning water cannot run onto the enclosure.
• Stock preparation: vibration from refiners and screens, plus steam and stock splashes. Specify seismic or vibration bracing rated for the refiner foundation, flexible bus connections, and a transformer located outside the splash zone.
• Effluent and water treatment: outdoor, hydrogen sulphide in the sewer atmosphere, and pumps that start frequently. Oil-immersed units with C4/C5 coating, sealed cable boxes, and a maintenance access path that does not require shutting the plant.

Ambient correction is not a rounding error. A 4 MVA cast-resin transformer rated at 40 °C ambient loses roughly 5–8% of its capacity at 45 °C, and a 33 kV oil-immersed unit loses about 1% per kelvin above its rated ambient. In a mill where the machine hall is 45 °C for six months of the year, that correction has to be in the enquiry, not discovered at commissioning. The same applies to altitude, which matters for mills in highland locations; the correction method is the one used for any distribution transformer, and it is worth checking the derating curve before fixing the rating. The relationship between cooling stage, temperature rise and permissible loading is covered in the cooling methods and temperature rise guide.

How should the mill be configured for reliability?

A paper mill runs 350 days a year and loses USD 30,000–90,000 per unplanned stop, so the redundancy question is economic rather than philosophical. The four configurations below cover almost every mill; the choice is usually made per area, not for the whole site.

For the paper machine specifically, the pattern that works is four transformers on two sections with a normally open tie and a fast transfer scheme: each pair carries half the machine at 80% loading, and on loss of one unit the tie closes and the machine continues at reduced speed with the coater and one winder shed. That is a deliberate compromise — full 2 × 100% on a 16 MVA machine bus is hard to justify, while a single radial is indefensible.

One reliability detail is easy to get wrong after a shutdown. When the mill is re-energised, every transformer on the board draws its magnetising inrush at the same moment, and a 4 MVA unit can pull 8–12 times rated current for 100–200 ms. If the paper machine drives are already enabled when the incomer closes, the voltage dip from the inrush is enough to trip them. The fix is a staged energisation sequence — transformers first, drives enabled 2–3 s later, sections accelerated in order — and the physics behind the current is explained in the transformer inrush current guide.

Why does a paper mill need a step-up transformer as well as step-down?

Most integrated mills generate a large part of their own power. A recovery boiler plus a biomass boiler plus an extraction-condensing turbine typically gives 30–60 MW at 11 kV, and the mill sells surplus to the grid in some markets and imports at night in others. That requires a generator step-up transformer at 11/132 kV or 11/110 kV, sized for the turbine rating with an overload margin, and it changes the protection and metering philosophy of the whole site.

Three consequences for the transformer specification:

• Reverse power flow: the mill’s 33 kV and 132 kV transformers must be specified for bidirectional power, with directional overcurrent and reverse-power relays, and with tap ranges that work in both directions. A transformer specified only for import will have an OLTC range that runs out of taps when the mill exports at 105% voltage.
• Harmonic flow to the grid: when the mill exports, the drive harmonics from the paper machine flow through the step-up transformer to the utility connection point. IEEE 519 limits apply at that point, not at the 690 V bus, and the utility will measure them. This is the reason a mill that is compliant at its own 11 kV board can still receive a power quality complaint from the network operator.
• Islanding: if the utility supply is lost, the mill may run islanded on its own generation. The step-up transformer then carries the full mill load with a very different fault level and no grid inertia behind it, and the load-shedding scheme has to be designed for that case rather than for the grid-connected one.

What does loss capitalisation look like for a mill running 8,000 hours a year?

A paper mill transformer runs continuously, so its losses are a capital cost, not an operating detail. Capitalise them at 8,000 hours a year, 0.08 USD/kWh, 20 years and a 6% discount rate, and the annuity factor is 11.47. One kilowatt of no-load loss is then worth 8,000 × 0.08 × 11.47 = USD 7,341, and one kilowatt of load loss at a 70% load factor is worth 8,000 × 0.49 × 0.08 × 11.47 = USD 3,597.

Apply it to a 4 MVA unit with 3.2 kW no-load loss and 32 kW load loss: 151,040 kWh a year, USD 12,083 a year, USD 138,593 over the 20-year life. That is the number to compare against the price difference between an S13 and an S20 grade unit, or between a standard and a low-loss design. In most cases the low-loss unit wins on a mill duty cycle and loses on a 2,000-hour duty cycle, which is why the same transformer can be the right choice for a paper mill and the wrong choice for a construction site. The full method, including how to handle load growth and how to treat the capitalised cost alongside the purchase price, is set out in the total cost of ownership calculation.

What belongs in the RFQ for a paper mill transformer package?

A paper mill enquiry fails in one of two ways: the supplier quotes a standard distribution transformer and the harmonics, the starting duty or the corrosion environment are discovered later; or the enquiry is so open that every offer is for a different product. The data list below closes both gaps. It is also the list to compare offers against, because a quote that cannot answer items 5, 6 and 9 is not a quote for this application.

Item 13 deserves its own survey. A 40 MVA, 132/33 kV unit weighs 60–80 t with oil and needs a route survey for bridge limits, overhead lines and turning radii before the order is placed; a 4 MVA cast-resin unit for the machine hall may have to pass through a 2.5 m door and be rigged into place before the roof is closed. The transport and installation constraints, including nitrogen-padded oil shipping and the tilt limits during handling, are set out in the transportation and installation guide.

Frequently asked questions

Can one 33/0.69 kV transformer feed the whole paper machine?

Technically yes, practically no. A single 16 MVA unit at 690 V would need 13,400 A of busbar, would put the entire machine behind one failure, and would concentrate the harmonic current of every drive on one winding. Four 4 MVA units on two sections cost more in switchgear but less in lost production, and they let you take one unit out for maintenance without stopping the machine.

Do I still need a K-rated transformer if all my drives are AFE?

No, and paying for one is a common waste. AFE drives with LCL filters produce 3–5% THDi, which corresponds to a K-factor close to 1. A standard transformer with a properly sized neutral is sufficient. The K-rating becomes relevant when 6-pulse drives remain on the bus — typically in a retrofit where part of the machine was upgraded and part was not.

What impedance should I specify for a transformer feeding refiners?

Choose it for fault current and regulation first, then check the start. On an 11 kV, 20 MVA transformer, 8% keeps the fault current at 13.1 kA and matches 16 kA switchgear; going to 10% saves one switchgear frame but adds 4% to the dip at a direct-on-line refiner start, which is the wrong trade. If the refiner starts direct-on-line and the dip exceeds 10%, the answer is a VFD or a soft starter, not a different transformer.

Can a paper mill transformer be installed indoors?

Yes, and most 690 V machine transformers are. Cast-resin dry-type units are designed for indoor installation, need no bund or fire wall, and are available with IP54 enclosures for dusty and humid areas. The constraints are ambient temperature, ventilation and dust management, not the transformer technology itself. Oil-immersed units above 20 MVA indoors require a fire-rated cell, a bund and usually a deluge system, which is why they are normally placed outdoors or in a dedicated substation building.

Sources / 资料来源

Source: IEC 60076-1 — Power transformers, general requirements.
Source: IEC 60076-2 — Temperature rise for liquid-immersed transformers.
Source: IEC 60076-11 — Dry-type power transformers.
Source: IEC 60038 — Standard voltages, including the 690 V level.
Source: IEC 60034-12 — Starting performance of single-speed three-phase cage induction motors.
Source: IEEE C57.110-2018 — Establishing transformer capability when supplying non-sinusoidal load currents.
Source: IEEE 519-2014 — Harmonic control in electric power systems.
Source: IEEE C57.12.00 — General requirements for liquid-immersed distribution, power and regulating transformers.
Source: GB/T 6451-2015 — Technical parameters and requirements for three-phase oil-immersed power transformers.
Source: GB 20052-2020 — Minimum allowable values of energy efficiency and energy efficiency grades for power transformers.
Source: GB/T 1094.1-2013 — Power transformers, general.
Source: Commission Regulation (EU) No 548/2014 — Ecodesign requirements for small, medium and large power transformers, Tier 2 from 1 July 2021.
Source: U.S. Department of Energy, 10 CFR Part 431 Subpart K — Distribution transformer energy conservation standards.
Source: ISO 12944-2 — Corrosivity categories for steel structures, category C5-M for coastal and industrial atmospheres.
Source: NFPA 70 (National Electrical Code) Article 450 — Transformers and transformer vaults.
Source: NFPA 13 — Installation of sprinkler systems, for the mill fire load context.
Source: FM Global Data Sheet 5-4 — Transformers, insurer requirements for separation and fire protection.
Source: IEC 60079-10-1 — Classification of hazardous areas, where solvent-based coating kitchens apply.
Source: QDTB engineering experience — paper mill area load lists, 690 V sectional drive bus sizing, refiner starting studies and harmonic derating practice. Values marked as industry experience are typical ranges from mill projects, not published standards.

If you are preparing an enquiry for a paper machine or pulp mill substation, the fastest route to a comparable offer is to send the area load list, the single-line diagram, the drive schedule with front-end types, and the largest motor-start data together. Our engineering team can turn that into a sized transformer specification, a harmonic derating calculation and an indicative budget in one pass. You can also start from the power system calculators to check the fault level, the voltage dip and the loss capitalisation before you approach any supplier — and if the question is which transformer type belongs where, the comparison between oil-immersed and dry-type transformers covers the decision in more detail.